8/6/2019 SNA-D-11-00271

1/22

Sensors & Actuators: A. PhysicalManuscript Draft

Manuscript Number: SNA-D-11-00271

Title: A NOVEL HUMID ELECTRONIC NOSE COMBINED WITH A ELECTRONIC TONGUE FOR ASSESSING

DETERIORATION OF WINE

Article Type: Research Paper

Keywords: Electronic tongue

Electronic nose

Wine quality

Abstract: We report herein the use of a combined system for the analysis of the spoilage of wine whenin contact with air. The system consists of a potentiometric electronic tongue and a humid electronic

nose. The potentiometric electronic tongue was built with thick-film serigraphic techniques usingcommercially available resistances and conductors for hybrid electronic circuits; i.e. Ag, Au, Cu, Ru,AgCl, and C. The humid electronic nose was designed in order to detect vapours that emanate from thewine and are apprehended by a moist environment. The humid nose was constructed using a piece of thin cloth sewn, damped with distilled water, forming five hollows of the right size to introduce theelectrodes. In this particular case four electrodes were used for the humid electronic nose: a glasselectrode, aluminium (Al), graphite and platinum (Pt) wires and an Ag-AgCl reference electrode. Thehumid electronic nose together with the potentiometric electronic tongue were used for the evaluationof the evolution with time of wine samples. Additionally to the analysis performed by the tongue andnose, the spoilage of the wines was followed via a simple determination of the titratable (total) acidity.

8/6/2019 SNA-D-11-00271

2/22

A NOVEL HUMID ELECTRONIC NOSE COMBINED WITH A ELECTRONIC TONGUE1FOR ASSESSING DETERIORATION OF WINE2

Luis Gil-Snchez a,b,* , Juan Soto a,c , Ramn Martnez-Mez a,c,d , Eduardo Garcia-Breijo a,b ,3Javier Ibez a,b , Eduard Llobet e 4

aCentro de Reconocimiento Molecular y Desarrollo Tecnolgico (IDM), Unidad Mixta Universidad 5Politcnica de Valencia Universidad de Valencia, Spain6

b Departamento de Ingeniera Electrnica. Universidad Politcnica de Valencia, Camino de Vera s/n.7 E-46022 Valencia, Spain8

c Departamento de Qumica. Universidad Politcnica de Valencia, Camino de Vera s/n. E-460229Valencia, Spain.10

d CIBER de Bioingeniera, Biomateriales y Nanomedicina (CIBER-BBN)11e MINOS-EMaSf, Department Enginyeria Electrnica, Universitat Rovira i Virgili (Tarragona - Spain)12

*Corresponding author: Luis Gil-Snchez, Phone: 34-963877608, Fax: 34-963877609, e-mail:13lgil@eln.upv.es 14

15

Abstract: We report herein the use of a combined system for the analysis of the spoilage of wine16when in contact with air. The system consists of a potentiometric electronic tongue and a humid17electronic nose. The potentiometric electronic tongue was built with thick-film serigraphic techniques18

using commercially available resistances and conductors for hybrid electronic circuits; i.e. Ag, Au,19Cu, Ru, AgCl, and C. The humid electronic nose was designed in order to detect vapours that emanate20from the wine and are apprehended by a moist environment. The humid nose was constructed using a21piece of thin cloth sewn, damped with distilled water, forming five hollows of the right size to22introduce the electrodes. In this particular case four electrodes were used for the humid electronic23nose: a glass electrode, aluminium (Al), graphite and platinum (Pt) wires and an Ag-AgCl reference24electrode. The humid electronic nose together with the potentiometric electronic tongue were used for25

the evaluation of the evolution with time of wine samples. Additionally to the analysis performed by26the tongue and nose, the spoilage of the wines was followed via a simple determination of the27titratable (total) acidity.28

Keywords: Electronic tongue, electronic nose, wine quality29

30

anuscriptk here to view linked References

mailto:lgil@eln.upv.esmailto:lgil@eln.upv.eshttp://ees.elsevier.com/sna/viewRCResults.aspx?pdf=1&docID=8617&rev=0&fileID=261898&msid={1D147800-B79B-4142-AC5A-78361F2772C0}http://ees.elsevier.com/sna/viewRCResults.aspx?pdf=1&docID=8617&rev=0&fileID=261898&msid={1D147800-B79B-4142-AC5A-78361F2772C0}mailto:lgil@eln.upv.es

8/6/2019 SNA-D-11-00271

3/22

INTRODUCTION31

The determination of wine quality has always been a dynamic field of research. The main constituents32of wine are water, ethanol and acids, but the elements that distinguish one wine from another, namely,33their quality, area of origin, vintage, etc are a large number of constituent components that are present34in very small proportions. Besides, the wine can be considered as a living product because its35chemical and organoleptic characteristics change over time. This makes the analysis of wine a rather36difficult task because, in addition to the determination of key components in small proportions,37analysis must be performed in different stages of its manufacture, storage and consumption. Another38reason for the interest in wine analysis resides in the quality control and detection of possible39adulterations or spoilage. Currently the basic technique to asses wine quality is the use of a panel of40trained experts that carry out sensory analysis based on colour, taste and flavour. However this is often41

time-consuming and requires skilled personnel. Therefore, there is an increasing interest in the42development of low-cost, easy-to-use analytical protocols able toin situ and at site monitoring wine43quality. In this context the use of electronic tongues and noses is a timely research field and a good44alternative to be used in the characterization of wines and other foodstuffs [1,2].45

Electronic noses or tongues are inspired in the mode mammalians recognize samples via46olfaction and taste senses. In this approach, sensors do not have to be selective but to respond47unspecifically to a group of related chemical species. The sensors are then integrated in an array and48their response is analysed by suitable pattern recognition procedures. These systems usually display49comparative/qualitative rather than quantitative information. Several research groups have developed50gas multisensor systems to determine the aroma of wine using electronic noses based on metal oxides,51such as SnO2 [3,4], WO3 [5], surface acoustic wave (SAW) devices [6] or gas chromatography mass52spectrometry GC MS [7] and compared their performance with sensory analysis [8]. In most cases53electronic noses commonly use resistive sensors, whose impedance varies with the presence of certain54gases. However it has been reported that the analysis of alcoholic beverages with semiconductor-55

based electronic noses is a difficult task due to the presence of high ethanol and water concentrations,56which trend to mask the presence of other important volatile compounds and contribute to the57shortening of sensor life [9]. In that sense, various solutions for dehydration and dealcoholisation of58wine samples have been proposed. [10].59

On the other hand, electronic tongues have also been used for the analysis of wines employing60arrays of electrochemical Langmuir-Blodgett film sensors and voltammetry [11], impedance61spectroscopy [12], amperometry [13] or potentiometry [14]. In relation to sensor technologies, thin-62film electrodes have been used for the analysis of white wines and artificial wines [15]. Additionally it63should be noted that some authors have studied the possibility of using a combination of electronic64

8/6/2019 SNA-D-11-00271

4/22

noses and tongues for wine analysis, [16,17] whereas more advanced systems combining electronic65nose, electronic tongue and an electronic eye have also been reported [18].66

Following our interest in the design of electronic tongues and noses [19] we report on the use67of a combined system for the analysis of the spoilage of wine when in contact with air. The system68

consists of a potentiometric electronic tongue and a humid electronic nose. The latter, in contrast to69impedance sensors, uses electrodes that are placed on a wet setting. This proposed system detects70vapours that emanate from the wine and are apprehended by a moist environment. The design of such71a humid electronic nose is, as far as we know, new and allows the use of classical electrochemical72techniques commonly employed in e-tongues for the analysis of volatile compounds. In this paper we73report the use of such a humid electronic nose together with a potentiometric electronic tongue for the74evaluation of the time evolution of wine samples.75

76

MEASUREMENT SYSTEM77

The measurement system used consists of two arrays of electrodes, for the humid electronic nose and78for the electronic tongue, that are attached to a common measurement equipment. The measurement79set is formed by a signal conditioning system, which comprises amplifiers with very high impedance80and filters for rejecting the 50 Hz noise, a system for data acquisition and a multivariate data analysis81

engine (Fig. 1). This measurement system is self-built. Details of its construction have been reported82elsewhere [20].83

84

Insert here Figure 185

86

Electronic Tongue Electrodes87

In order to make the electronic tongue, a potentiometric measurement system has been developed,88which consists of electrodes built employing thick-film serigraphic techniques. The pastes are89commercially available and commonly used in the manufacture of thick-film resistances and90conductors for hybrid electronic circuits. Each one of these pastes has an active element, which we91employed to obtain potentiometric electrodes. The active elements are: Ag, Au, Cu, Ru, AgCl, and C.92Although in the electronic tongue these materials are used as non-specific electrodes, their properties93as suitable surfaces sensitive to chemical species have already been studied [21]. Some materials have94been duplicated on the multisensor system, finally obtaining 9 potentiometric electrodes. Fig. 2 shows95

8/6/2019 SNA-D-11-00271

5/22

the electrode distribution on the multisensor board and the tracks and pads for connecting the board96using a flat wire. More details can be found elsewhere [22].97

98 Insert here Figure 299

100The Humid Electronic Nose101

The humid electronic nose was inspired in biological olfaction which needs a wet medium for the102absorption of volatile components. The humid electronic nose was constructed using a piece of thin103cloth, sewn forming five hollow spaces of the appropriate size to introduce five different wire104electrodes (i.e. four active electrodes and one reference electrode). Electrodes are tightly fit so a105perfect contact between the electrode and the cloth is accomplished. In order to achieve a good106conductivity between the active electrodes and reference electrode the cloth was damped with distilled107water. Appropriate results were obtained using nylon textiles. Fig. 3 shows a scheme of the humid108nose. Volatile compounds in wine will evaporate and will be dissolved in the damp fabric were the109potentiometric electrodes are included. In this particular case four electrodes were used for the humid110electronic nose: a glass electrode, aluminium (Al), graphite and platinum (Pt) wires and an Ag-AgCl111reference electrode.112

113 Insert here Figure 3114

115

EXPERIMENTS116

Samples117

Three Spanish table wines were analyzed with the humid electronic nose and the electronic tongue118

combined system; i.e. two red and one white wine. The red wines were labelled RA and RB and the119white wine was labelled W. All the wine bottles were opened the same day, and measurements were120made at days 1, 5, 9, 15, 19, 22, 28, 36 and 48. During these days the bottles remained opened. Wines121were stored at a constant temperature of 15C. Each working day the wines were measured using the122humid electronic nose and the electronic tongue. Also each working day the total titratable acidity was123measured for all three wines.124

Acidity Determination125

8/6/2019 SNA-D-11-00271

6/22

Titratable (total) acidity in wines was determined by simple acid base titration using NaOH with126either visual or potentiometric detection.127

Measurement using the electronic tongue and the humid electronic nose128

To perform the measurements with the electronic tongue system the multisensor board (see Fig. 2)129containing an Ag-AgCl reference electrode, was dipped into a vessel containing the corresponding130wine sample. Then the potential between each active electrode and the reference electrode was131measured. Measurements were performed automatically by sampling every five seconds during a time132of about ten minutes to achieve electrochemical stability.133

To make measurements with the humid electronic nose the corresponding wine was poured134into a container that comprised the set of electrodes wrapped by a damp fabric and the system was135

sealed. Thus, volatiles emanating from the wine permeated the wet cloth, changing the characteristics136of the electrochemical cells formed by an active electrode and the reference electrode. Similar to the137electronic tongue system, the potential between each electrode and the reference electrode was138measured during ten minutes in order to reach electrochemical stability.139

For both the electronic nose and the humid electronic nose a standard solution was prepared.140Before measurements with the wine samples the set of electrodes were dipped in buffered water at pH1415 (standard solution) and the potential for each electrode was measured. This was set as a cero142

reference value. The data for the posterior analysis with the wine samples was taken as the difference143between the cero reference and the measured potential for each sample.144

The final data array contained of 27 rows (3 wines x 9 days) and 13 columns (electrodes from145the humid electronic nose (4) and from the electronic tongue (9)).146

147

RESULTS148

As stated above, following our interest in the design of electronic tongues and noses for different149applications, we attempted to apply the use of these devices to the analysis of the spoilage of wine150when in contact with air. In the first step a potentiometric electronic tongue was used. The electronic151tongue consisted of a set of electrodes built with thick-film serigraphic techniques. The active152elements in the electrodes were: Ag, Au, Cu, Ru, AgCl, and C (see experimental section). Despite the153fact that similar electronic tongues have been used by some of us in certain applications such fish154freshness analysis, [23] in this case, the tongue barely was able to discriminate the evolution of the155

samples through time (data not shown) in the wine spoilage process. In this particular case, this low156discrimination using the electronic tongue may be attributed to the possible masking ability of water157and ethanol that are found in very high concentrations in wine samples. Therefore it occurred to us158

8/6/2019 SNA-D-11-00271

7/22

that a combination of tongue and nose might be more suitable for monitoring the evolution of wine159through time due to the presence, after some time, of volatile derivatives usually associated to wine160spoilage (vide infra). Additionally we were especially interested in the design of a humid electronic161tongue as a suitable alternative to the use of classical electronic noses based on metal oxides. One162advantage of this humid electronic tongue, that was designed to evaluate the presence of volatile163compounds, is the possibility of using typical electrochemical electrodes such as those usually found164in electronic tongues. The nose involves the use of potentiometric electrodes (wires of Al, C165(graphite), Pt and a glass electrode) and a reference electrode (Ag-AgCl) that are placed in a damp166cloth sewn. We used such humid electronic nose in combination with a potentiometric electronic167tongue for monitoring the evolution of wine with time. As sated above for this set of experiments168three wines were used; i.e. two red wines (RA and RB) and one white wine (W). The wine bottles169were opened the same day and measurements were made at days 1, 5, 9, 15, 19, 22, 28, 36 and 48170after they were opened.171

172

Insert here Figure 4173

174Wine is a complex mixture of compounds that usually derive from three major sources; i.e.175

grapes, microorganism and (if used) oak. Additionally wine spoilage is a rather complex process in176which acetic acid bacteria play an important role because their metabolites result in disagreeable wine177sensory characteristics. The concentration of acetic acid in wine is an important parameter because it178is the head of the deterioration process of the wine, causing the spoilage effect that set the wine a179vinegary taste. Acetic acid is the major volatile acid in wine and the main constituent of wine volatile180acidity and is usually considered undesirable in wines. At high concentrations, acetic acid gives a sour181flavour and a vinegar-like aroma. Acetic acid is produced by bacteria through the metabolism form182ethanol to acetaldehyde to acetic acid. Both volatile intermediate metabolites acetaldehyde and ethyl183ester of acetic acid also contribute to the sensory spoilage of wine. [24,25] Whereas in the spoilage of184wine process other acids such as citric acid and malic acid remain at the same concentrations, the185amount of acetic acid, ethyl acetate and acetaldehyde increases significantly. Therefore, as a simple186control of wine spoilage as a function of time we determined the titratable (total) acidity in wines RA,187RB and W by simple acid base titrations at days 1, 5, 9, 15, 19, 22, 28, 36 and 48. The results are188shown in Figure 4. The acidity measurements reveal that, for all wine types, little variation occurs in189the first days and a sharp increase in acidity is observed in the last days. Note also that the variation of190

acidity slightly depends on the type of wine.191

8/6/2019 SNA-D-11-00271

8/22

192Result combining the electronic tongue and the electronic nose193

A suitable mode to interpret the results obtained from the use of the electronic tongue and the humid194electronic nose on the different wine samples is to use multivariate analysis which is a statistical195

method able to determine the contribution of various factors on a single event. To perform this type of 196analysis Matlab 2010 with the appropriate Toolbox (PLS_Toolbox and Statistics Toolbox) has been197used. One of the methods most commonly used of multivariate analysis is Principal Component198Analysis (PCA). For this purpose, new orthogonal directions in the variable space are searched, called199Principal Components (PC). The representation of the first two principal components (PC1 and PC2)200in a two-dimensional graphic often provides a fairly significant result of the actions and behavior of 201the system. [26]202

The first two principal components of a PCA of data from the combined electronic tongue and203humid electronic nose system are represented in Figure 5. Samples of three wines RA, RB and W of a204given day appear close to each other. While the content of acetic acid in wine increases slightly (i.e.,205from day 1 to day 36), the scores of the wines tested move from right to left along the first principal206component. At day 48, a strong variation in the second principal component is observed. This may be207due to the sharp increase in the concentration of acetic acid 48 days after the bottles were opened but208also to the occurrence of other compounds associated to wine degradation. It is interesting to note that209

the system is capable of distinguishing the measurements of the first days, despite the fact that acidity210measurements of those days were almost identical. That is, the proposed system appears to be able to211detect components that do not directly affect the acidity of wine. One of the main problems when212using metallic electrodes is the possible occurrence of response or baseline drift that could lead to213obtain irreproducible data or to mistakenly believe that drift variations are changes in wine spoilage.214In order to control these possible drifts, prior to the measurement of the wine samples, the electrodes215were immersed in an aqueous sample (25oC buffered at pH 5, standard solution) and the data for the216

posterior analysis was taken as the difference between the cero reference and the measured potential217for each wine sample.218

219 Insert here Figure 5220

221

In order to understand the relationship between the diverse electrodes used, the loads of each222

electrode in the first two Principal Components are shown in Fig. 6. This figure helps to verify the223different significance of the electrodes in the electronic nose respect the ones used in the electronic224

8/6/2019 SNA-D-11-00271

9/22

tongue. The loads of the electronic tongue electrodes are in the left-hand side of the plot (loads on225PC1 are negative), all of them with a homogenous activity.226

Glass, Aluminium and Platinum electrodes of the humid electronic nose show very different227loads with respect of the electronic tongue ones, that is to say, the former electrodes are independent228

of the latter. These electrodes, which appear in the upper side of Fig. 6, are responsible for the position229of measurements performed the last day (48), which appear apart in the score plot (Fig. 5). A possible230chemical interpretation of these results is that the last day (48) the wines give off aromas with a higher231acidity than in previous days because wines are soured.232

Fig. 6 also shows that the different electrodes significantly contribute to the first two principal233components. Only the loads of the graphite electrode are close to zero. In fact this electrode had an234erratic behaviour and could be removed from the set of electrodes without significant changes in the235PCA scores plot.236

Insert here Figure 6 237

238By selecting the first two principal components in the principal component analysis, only239

69.38% of the data variance (i.e. information) provided by the electronic tongue and humid electronic240nose are actually used. This value is not very high, which means that there is not a high degree of 241

correlation between the different electrodes employed. This fact is quite logical due to the different242nature of the electrodes employed in the electronic tongue and humid electronic nose.243

To achieve an analysis in which all the information (i.e. all data variance) from the electronic244tongue and the humid electronic nose is used, a cluster analysis with dendrograms has been245performed. A dendrogram acts as genealogical tree of the measurements. The measurements are246associated to each other to complete the tree. To obtain a dendrogram different algorithms can be247employed but the most commonly used is K-means clustering. This algorithm aims to partitionn 248observations intok clusters in which each observation belongs to the cluster with the nearest mean249[27]. Fig, 7 shows the dendogram obtained when a k-means clustering is applied to the wine data.250Measurements performed at day 48 are clearly different from any other measurement performed251before and appear grouped in a single cluster. Two other clusters can be easily inferred, which group252together measurements performed between days 1 to 9 and from days 15 to 36, respectively. If we253consider now the cluster grouping measurements from day 1 to day 9, measurements of the first day254are clearly different from the others suggesting that early stages of wine evolution can be detected. For255

the cluster that groups measurements performed between days 15 and 36, the initial and last256

8/6/2019 SNA-D-11-00271

10/22

measurements within this group (i.e. those of day 15 and day 36) appear clearly discriminated in the257dendogram. These results are consistent with those obtained in the PCA plot.258

259

Insert here Figure 7 260

CONCLUSIONS261

In summary we have combined a potentiometric electronic tongue with a humid electronic nose and262have applied the joint electronic system to the analysis of the spoilage of wine. The potentiometric263electronic tongue was designed using thick-film serigraphic techniques and commercially available264pastes originally intended for resistances and conductors (i.e. Ag, Au, Cu, Ru, AgCl, and C). The265designed humid electronic nose contained a glass electrode, aluminium (Al), graphite and platinum266(Pt) wires and a Ag-AgCl reference electrode that were included in a damp fabric. The humid267electronic nose responds to volatile compounds that evaporate from the wine samples and that268dissolved and concentrated in the wet cloth were the potentiometric electrodes are included. The269process of wine spoilage when in contact with air was followed with the combined electronic system270with good results. Additionally, the spoilage was also assessed by using a simple determination of the271titratable (total) acidity. To check the system, three Spanish table wines, two red and one white, have272been analyzed. The wine bottles were all opened at the same time and measurements were made with273the wines for 48 days. A clear discrimination as a function of time was achieved.274

275

AKNOWLEDGEMENTS276

We thank the Spanish Government (project MAT2009-14564-C04) the Generalitat Valenciana277(project PROMETEO/2009/016) for support. We would also like to thank the Universidad Politcnica278de Valencia for support (Primeros Proyectos de Investigacin PAID-06-09).279

280

REFERENCES281

[1] R. S. Jackson. Sensory Perception and Wine Assessment. Wine Science. Third Edition (2008)282641-685283

[2] J. Zeravik, A. Hlavacek, K. Lacina, P. Skldal. State of the Art in the Field of Electronic and284Bioelectronic Tongues Towards the Analysis of Wines. Electroanalysis 21 (2009) 2509 2520285

8/6/2019 SNA-D-11-00271

11/22

8/6/2019 SNA-D-11-00271

12/22

[14] A. Legin, A. Rudnitskaya, L. Lvova, Yu.Vlasov, C. Di Natale, A. DAmico. Evaluation of 317Italian wine by the electronic tongue: recognition, quantitative analysis and correlation with318human sensory perception Anal. Chim. Acta 484 (2003) 33 44319

[15] G. Verrelli, L. Francioso, R. Paolesse, P. Siciliano, C. Di Natale, A. DAmico, A. Logrieco.320

Development of silicon-based potentiometric sensors: Towards a miniaturized electronic tongue.321Sens. Actuators, B, 123 (2007) 191 197322

[16] L. Rong, W. Ping, H, Wenlei. A novel method for wine analysis based on sensor fusion technique323Sens. Actuators, B, 66 (2000) 246-250324

[17] S. Buratti, S. Benedetti, M. Scampicchio, E.C. Pangerod. Characterization and classification of325Italian Barbera wines by using an electronic nose and an amperometric electronic tongue. Anal.326

Chim. Acta, 525 (2004) 133 139327

[18] M.L Rodrguez-Mndez, A. Arrieta, V. Parra, A. Bernal, A. Vegas, S. Villanueva, R. Gutirrez-328Osuna; Combination of an electronic nose, an electronic tongue and an electronic eye for the329Analysis of Red Wines aged with alternative methods. IEEE Sensors Journal, 4, (2004), 348-354330

[19] R. H. Labrador, R. Masot, M. Alcaiz, D. Baigts, J. Soto, R. Martnez-Maez, E. Garca-Breijo,331L. Gil, J.M. Barat. Prediction of NaCl, nitrate and nitrite contents in minced meat by using a332voltammetric electronic tongue and an impedimetric sensor Food Chemistry 122 (2010) 864 870333

[20] R. Martinez-Maez, J. Soto, E. Garcia-Breijo, L. Gil, J. Ibaez, E.Llobet, An electronic tongue334design for the qualitative analysis of natural waters. Sens. Actuators, B, 104 (2005) 302-307335

[21] J. Soto, R. H. Labrador, M.D. Marcos, R. Martnez-Mez, C. Coll, E. Garca-Breijo, L. Gil. A336model for the assessment of interfering processes in Faradic electrodes. Sens. Actuators, A, 142337(2008) 56-60338

[22] R. Martnez-Mez, J. Soto, E. Garca-Breijo, L. Gil, J. Ibez, E. Gadea. A multisensor in thick-339film technology for water quality control. Sens. Actuators, A, 120 (2005) 589-595340

[23] L. Gil, J. M. Barat, E. Garcia-Breijo, J. Ibaez, R. Martnez-Mez, J. Soto, E. Llobet, J.341Brezmes, M. C. Aristoy, F. Toldr. Fish freshness analysis using metallic potentiometric342electrodes. Sens. Actuators, B, 131 (2008) 362 370343

[24] V. Loureiro, M. Malfeito-Ferreira. Spoilage yeasts in the wine industry. International Journal of 344

Food Microbiology, 86 (2003) 23 50345

8/6/2019 SNA-D-11-00271

13/22

[25] A. Gonzlez, N. Hierro, M. Poblet, A. Mas, J. M. Guillamn. Application of molecular methods346to demonstrate species and strain evolution of acetic acid bacteria population during wine347production. International Journal of Food Microbiology, 102 (2005) 295-304348

[26] B.D. Ripley, Pattern Recognition and Neural Networks. Cambridge U.K: Cambridge Univer.349

Press, 1996.350

[27] R. Gutierrez-Osuna. Pattern Analysis for Machine Olfaction: A Review. IEEE SENSORS351JOURNAL, 2 (2002) 189-202352

353

354

8/6/2019 SNA-D-11-00271

14/22

VITAE355

Luis Gil-Snchez was graduated in electronic engineering from the Universitat de Valencia (Spain) in3561998, and received his PhD in 2007 from the Universidad Politcnica de Valencia (UPV). He is357assistant professor of electronic technology in the Electronic Engineering Department at UPV. He is a358member of the Institute of Molecular Recognition and Technological Development (IDM). His main359areas of interest are the chemical sensors, instrumentation systems and pattern recognition for360electronic tongues.361

Juan Soto was graduated in Chemistry from the Universitat de Valencia in 1981 and received his PhD362in 1986 in the same University. He is currently associate professor in the Department of Chemistry at363the Universidad Politcnica de Valencia. His main areas of interest are the development of chemical364chemosensors and probes, especially those based on electrochemical processes.365

Ramn Martnez-Mez was graduated in Chemistry from the Universitat de Valencia in 1986 and366received his PhD in 1990 in the same University. After a postdoctoral period at Cambridge (UK), he367 joined the Department of Chemistry at the Universidad Politcnica de Valencia. He became full368professor in 2002. His main areas of interest are in the field of chromo-fluorogenic chemosensors and369molecular probes for anions, cations and neutral chemical species.370

Eduardo Garca Breijo was graduated in electronic engineering from the Universitat de Valncia371(Spain) in 1997, and received his PhD in 2004 from the Universidad Politcnica de Valencia (UPV).372He is an assistant professor of Electronic Technologic of UPV. He is a member of the Institute of 373Molecular Recognition and Technological Development (IDM). His main areas of interest are the374development of multisensors in thick-film technology. 375

Javier Ibez Civera is Matrisse in Power Electronic and Control from the Universite Pierre et Marie376Curie (Paris VI) in 1994 and received his PhD in 2009 from the Universidad Politcnica de Valencia377

(UPV). He is assistant professor of Electronic Technology in the Electronic Engineering Department378 of the Universidad Politcnica de Valencia (UPV). He is a member of the Institute of Molecular379Recognition and Technological Development (IDM). His main areas of interest are water organic380contamination devices.381

Eduard Llobet received the degree in telecommunication engineering and the Ph.D. degree from the382Universitat Politcnica de Catalunya, Barcelona, Spain, in 1991 and 1997, respectively. He is383currently an associate professor in the Electronic Engineering Department, Universitat Rovira i384

Virgili, Tarragona, Spain. His main areas of interest are in the fabrication and modelling of 385

8/6/2019 SNA-D-11-00271

15/22

semiconductor chemical sensors and in the application of intelligent systems to complex odour386analysis.387

8/6/2019 SNA-D-11-00271

16/22

Electronic Tongue

HighImpedance

9

4

HighImpedance

Multiplexor+ Filter

PBFilter

DataAdquisition

Board PC1

PC2

.....x

x

x

x

x

++

++++

PC2

.....x

x

x

x

x

++

++++

Electronic Nose

Signal ConditoningSystem

Fig. 1. Block diagram of the measurement system

ure 1

8/6/2019 SNA-D-11-00271

17/22

Cu1

Cu2

C2

C1

Au1Au2

Ru

Ag

ClAg

Cu1

Cu2

C2

C1

Au1Au2

Ru

Ag

ClAg

Fig. 1. Potentiometric sensors for electronic tongue

ure 2

8/6/2019 SNA-D-11-00271

18/22

Wine

Moistened Bag

VGlass

Graphite Reference PtAl Glass

VGraph.

VPtVAl

Electrodes incontact with the bagand separatedbetween them

Fig. 1. Scheme of the humid electronic nose constructed using a set of electrodesclosed in a piece of thin moistened cloth

ure 3

8/6/2019 SNA-D-11-00271

19/22

Fig. 1. Variation of acidity in wines RA, RB and W

ure 4

8/6/2019 SNA-D-11-00271

20/22

Fig. 1. PCA scores of electronic tongue and nose combination. The number isrelated with the day the wines RA, RB and W wines were measured.

ure 5

8/6/2019 SNA-D-11-00271

21/22

Fig. 1. PCA loads of electronic tongue and nose combination

ure 6

8/6/2019 SNA-D-11-00271

22/22

48

1-9

15-36

48

1-9

15-36

Fig. 1. Dendrogram using K-means

ure 7